October 25, 2014

Surpluses as Far as the Eye Can See? It All Comes Down to Assumptions

Is the federal budget heading for unsustainable deficits or unsustainable surpluses?

It all depends on the long-term assumptions. 

Last week, the Government Accountability Office (GAO) issued an update of its long-term fiscal outlook for the federal government. As in prior reports,  GAO found that an extension of current law (the Baseline Extended simulation) leads to rising and eventually unsustainable debt “driven by a fundamental imbalance between revenue and spending, which, on the spending side, is driven by the aging of the population and rising health care costs.”

On the other hand, the President's Office of Management and Budget (OMB) released a new estimate last week showing that an extension of President Obama's budget policies would not just be sustainable but would lead to growing surpluses that would eventually pay off the national debt.

Not that OMB thinks this will actually happen. In fact, OMB calls the end result “unrealistic and undesirable.”

As explained by OMB in the Analytical Perspectives of the 2014 Budget, “These projections are not intended to be a prediction of future legislative action, nor are they intended to reflect explicit policy proposals for the years beyond 2023; rather, they are a mechanical extrapolation of the Budget policies.”

But in case the point is missed, OMB adds that the “base forecast shows that under 2014 Budget policies, in the long run the budget does not run deficits or increase the debt.”

It is tempting to dismiss this as a transparent attempt to portray the long-term effect of the President’s budget in the best possible light. However, it raises an important question: How will policymakers know if they have achieved a sustainable fiscal plan?

Any such assessment will depend on several important but highly uncertain assumptions about policy trends beyond the 10-year window used to score specific budget policies. The degree to which these assumptions can lead to remarkably divergent results over time is demonstrated by the GAO and OMB projections.

The GAO simulation begins with the February 2013 Congressional Budget Office (CBO) baseline (2014-2023).  For later decades, the GAO simulation assumes that Social Security and the major federal health care programs (Medicare, Medicaid, Children’s Health Insurance Program and the new exchange subsidies) will continue to grow faster than the economy (GDP). Revenues and all other spending are assumed to remain constant as a share of GDP after 2023.

Under these assumptions, debt held by the public would grow to nearly 100 percent of GDP in 2030 and to nearly 200 percent of GDP in 2050. By 2085, the debt would be 480 percent of GDP.

These results are not surprising.  Government agencies and outside observers using similar assumptions have been warning for years that the budget is on an unsustainable track.

What did come as a surprise was OMB’s “base case” projection that budget policy is sustainable. Starting with the President’s proposed budget through 2023, which is not very different than the CBO baseline (GAO’s starting point), it shows debt held by the public staying relatively flat as a share of GDP through 2040 (at about 76 percent) and then beginning a steady decline, which by 2074 would eliminate the debt. Beyond then, the government would accumulate assets that by 2085 would reach 57 percent of GDP.

If the purpose of such long-range projections is to give policymakers a reasonable estimate of where fiscal trends are headed, then the GAO projection is, on balance, more useful.

It all comes down to the underlying assumptions about health care costs, discretionary spending and revenue growth. On all three, OMB has a more optimistic view of the fiscal future. Health care costs are permanently restrained, discretionary spending shrivels to a shadow of its historic average and revenues continue to outpace the economy.

This optimism is a change from the past, when OMB’s projections were closer to GAO’s. What has changed is not so much substance but methodology.

One significant change is OMB’s new assumption that discretionary spending will grow with inflation and population beyond the 10-year budget window rather than at the same pace as the economy (GAO’s assumption). OMB’s assumption contributes a huge amount of deficit reduction over time but leaves discretionary spending at an implausibly low level.

Over the past 40 years discretionary spending has averaged 8.7 percent of GDP. It is currently at 7.6 percent of GDP and is projected to drop further as the tight budget caps and sequestration of the 2011 Budget Control Act take effect. Thus, GAO’s projection that discretionary spending will lock in at 5.5 percent of GDP over the long term (the 2023 baseline level) already signals a substantial and permanent departure from the norm. 

Yet OMB takes this a step further and assumes that discretionary spending will continue to decline relative to the economy, reaching 2.3 percent of GDP by 2085 – less than half of GAO’s projection and roughly one-quarter of the 40-year average. Last year, OMB rightly dismissed such an assumption as “not realistic.”

Another change in OMB’s methodology is to assume that revenues will be allowed to grow faster as a share of the economy because of “bracket creep” in which taxpayers are forced into higher tax brackets due to their incomes rising faster than inflation. This is different not just from GAO’s flat projection based on the current baseline beyond 2023, (19.1 percent of GDP), but from OMB’s own previous assumption that bracket creep would be mitigated by future legislative action.

For example, in last year’s long-term scenario OMB projected that revenues would rise to 20.9 percent of GDP in 2085, whereas this year’s projection takes revenues up to 23.8 by that year.

Unlike the discretionary spending assumption, this change is probably more realistic than GAO’s flat-line assumption. As OMB points out, “Allowing revenues to rise is methodologically consistent with the approach for the projections of Social Security, Medicare and other mandatory spending programs in that it projects the levels of revenues that would result under an extrapolation of the Budget policies.”

It does, however, represent a change from the traditional view that revenues will somehow be constrained as incomes grow, which is the rationale behind GAO’s method. Even as incomes have grown in the past, revenues have rarely exceeded 19 percent of GDP. That doesn’t mean they shouldn’t exceed that level or won’t in the future. It only means that assuming a higher level would be consistent with current law but inconsistent with experience.   

Finally, the new OMB projections are very optimistic regarding the extent to which Medicare and other health care costs will be controlled beyond the 10-year budget window.

The GAO baseline extended simulation assumes that the cost-control mechanisms of the Affordable Care Act (ACA, or “Obamacare”) will be fully implemented and effective. However, as GAO points out, “the [Medicare] Trustees, CBO and the CMS Actuary have questioned whether certain cost-containment mechanisms can be sustained over the long-term.”

OMB’s lower projection of Medicare spending also assumes the ACA’s cost containment mechanisms are fully implemented and effective (without GAO’s caveats) and then assumes the successful addition of the President’s proposed Medicare changes --  such as higher premiums for upper-income beneficiaries and a more aggressive cost-control target (GDP plus 0.5 percent) for the Independent Payment Advisory Board (IPAB), which would save money over time.

It is impossible to know how much of their lower Medicare projections are due to their baseline assumptions versus the effect of their new proposals because their analysis does not have a breakdown.

However, with regard to IPAB, it should be noted that no one has yet been appointed to serve on this board, prospects for a fully functioning board any time soon look bleak given the need for all 15 appointments to pass through a 60-vote Senate, and that even if there was a board it would have only limited tools to meet its ambitious cost-saving target. Over-reliance on IPAB at this point may thus be misplaced.

Policymakers negotiating a fiscal sustainability plan will need to agree in advance what the methodology will be used for projecting future costs. Because all long-range assumptions are highly uncertain, no one can say which are objectively “correct.” As we found out in recent days, with the right assumptions just about anything is possible.

For planning purposes, however, it would be prudent not to base policy changes on a best-case scenario. It’s always better to be pleasantly surprised than to find out that hard-fought political concessions on both sides are still not enough.